Struct EvalCriterion 

pub struct EvalCriterion { /* private fields */ }

An evaluation criterion applied to agent output.

Implementations

impl EvalCriterion

pub fn name(&self) -> &str

Name of this criterion.

pub fn score(&self, output: &str, expected: &str) -> f64

Synchronously score the output against expected (0.0–1.0). LLM-judge criteria cannot run on the sync path and return 1.0 here — use EvalCriterion::score_async for those.

pub async fn score_async(&self, output: &str, expected: &str) -> f64

Score the output, running an LLM judge if this is a judge criterion. A judge criterion scores 1.0 when no violation is flagged, else 0.0.

Trait Implementations

impl BitOr<EvalCriterion> for EvalComposite

type Output = EvalComposite

The resulting type after applying the | operator.

fn bitor(self, rhs: EvalCriterion) -> Self::Output

Performs the | operation.

impl BitOr for EvalCriterion

Compose two criteria with |.

type Output = EvalComposite

The resulting type after applying the | operator.

fn bitor(self, rhs: EvalCriterion) -> Self::Output

Performs the | operation.

impl Clone for EvalCriterion

fn clone(&self) -> EvalCriterion

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for EvalCriterion

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl From<EvalCriterion> for EvalComposite

A single criterion is a one-element composite, so E::suite().criteria(E::response_match()) works without an explicit |.

fn from(criterion: EvalCriterion) -> Self

Converts to this type from the input type.